Mirror an image using the CImg?

The title is pretty self-explanatory, but I will stress that I am a beginner at c++ and especially when it comes down to the CImg library. Here's the code I have so far and I tried experimenting with the flipped_idx (which didn't work when incorporated into the code) but I don't think I'm heading in the right direction. Or maybe I am, any help would be greatly appreciated!

I'm inputting the first two images below and merging them into one image.

Here's the merged image that I am trying to mirror on the y-axis.

#include <iostream>
#include "CImg.h"

using namespace std;
using namespace cimg_library;

const int imgscale = 400;

int main() {

CImg<unsigned char> player("input/player.bmp");

CImgDisplay green_disp(player, "Original 'player.bmp'");
green_disp.resize(imgscale, imgscale, true);
green_disp.move(10, 50);


CImg<unsigned char> bg("input/forest.bmp");

CImgDisplay bg_disp(bg, "Original 'forest.bmp'");
bg_disp.resize(imgscale, imgscale, true);
bg_disp.move(20 + imgscale, 50);

CImg<unsigned char> merged(player.width(), player.height(),1,3,0);
int w = player.width();
int h = player.height();
int imglen = w*h;
const CImg<float>
    flipped_idx = merged.get_mirror('y');


for (int i = 0; i < h; ++i) {
    for (int j = 0; j < w; ++j) {
        int idx = j + i*w;


        unsigned char colorR = player[idx];
        unsigned char colorG = player[idx + imglen];
        unsigned char colorB = player[idx + imglen * 2];

        if (colorG > 215 && colorR < 255) {
            //Background
            merged[idx] = bg[idx]; //red
            merged[idx + imglen] = bg[idx + imglen]; //green
            merged[idx + imglen * 2] = bg[idx + imglen * 2]; //blue
        }
        else {
            //Foreground
            merged[idx] = player[idx]; //red
            merged[idx + imglen] = player[idx + imglen]; //green
            merged[idx + imglen * 2] = player[idx + imglen * 2]; //blue
        }

    }
}

CImgDisplay merged_disp(merged, "Merged image");
merged_disp.resize(imgscale, imgscale,true); //change size
merged_disp.move(30 + imgscale*2, 50); //place it nicely on screen.
merged.save("output/merged.bmp");

If I try and stick to your code as closely as possible, but discard all the display stuff, I get this:

#include <iostream>
#include "CImg.h"

using namespace std;
using namespace cimg_library;

const int imgscale = 400;

int main() {

CImg<unsigned char> player("player.bmp");
CImg<unsigned char> bg("forest.bmp");

CImg<unsigned char> merged(player.width(), player.height(),1,3,0);
int w = player.width();
int h = player.height();
int imglen = w*h;
CImg<unsigned char> result;

for (int i = 0; i < h; ++i) {
    for (int j = 0; j < w; ++j) {
        int idx = j + i*w;

        unsigned char colorR = player[idx];
        unsigned char colorG = player[idx + imglen];
        unsigned char colorB = player[idx + imglen * 2];

        if (colorG > 215 && colorR < 255) {
            //Background
            merged[idx] = bg[idx]; //red
            merged[idx + imglen] = bg[idx + imglen]; //green
            merged[idx + imglen * 2] = bg[idx + imglen * 2]; //blue
        }
        else {
            //Foreground
            merged[idx] = player[idx]; //red
            merged[idx + imglen] = player[idx + imglen]; //green
            merged[idx + imglen * 2] = player[idx + imglen * 2]; //blue
        }
    }
}

result = merged.get_mirror('y');
result.save("merged.bmp");
}

If I was coding it, I would probably go with something more like this:

#include <iostream>
#include "CImg.h"

using namespace std;
using namespace cimg_library;

int main() {

   CImg<unsigned char> player("player.bmp");
   CImg<unsigned char> bg("forest.bmp");
   CImg<unsigned char> merged(player.width(), player.height(),1,3,0);
   CImg<unsigned char> result;

   // Extract Green channel and make into mask by thresholding
   CImg<unsigned char> mask = player.get_channel(1).threshold(215);
   mask.save("mask.bmp");

   // Select background or foreground according to mask
   merged = bg.mul(mask) + player.mul(1-mask);
   merged.save("merged.bmp");

   // Mirror and save
   result = merged.get_mirror('y');
   result.save("result.bmp");
}

Just FYI, the mask looks like this: